Hidden risks of using AI in construction

Artificial intelligence can save a construction project manager hours every day. It can also create risks that many companies haven't fully considered.

The biggest danger isn't that AI makes mistakes. It's that people stop questioning its answers.

The main risks of using AI in construction include:

1.    Project Data security and confidentiality

Construction teams regularly handle sensitive information, including contracts, drawings, tender documents, schedules and client data.

Uploading these documents to public or unapproved AI platforms may breach confidentiality agreements, intellectual property obligations and internal company policies

Companies should therefore provide approved, secure AI tools and establish clear rules regarding what employees can upload.

2.    Critical thinking skill degradation

Construction decisions depend on more than the information written in a prompt.

Site conditions, weather, sequencing, team experience, subcontractor relationships and unexpected constraints can completely change the correct course of action. AI rarely sees the whole project. It only sees the context it has been given.

Overreliance on AI weakens critical-thinking skills, particularly among junior employees who still need to develop practical judgement.

3.    Loss of competitive advantage

When every contractor uses the same tools to prepare proposals, presentations and tender responses, the results can quickly begin to sound identical.

Generic language can dilute the experience, specialist knowledge and lessons learned that make one company different from another. AI-generated content should therefore be shaped by real project experience and reviewed by people who understand the organization’s methods, values and strengths.

 . 

4.    Cost underestimation caused by blind trust in AI

Construction professionals can easily assume that a well-written AI cost estimate is also an accurate one.

For example, AI can generate a construction cost estimate that appears detailed and professionally presented, yet significantly underestimates the actual project cost. It may overlook industry standards, site-specific constraints, logistics, productivity factors, market conditions or lessons learned from previous projects.

Without benchmarking against historical project data and validation by experienced professionals, these estimates can be misleading.

5.    Safety risks from unverified AI recommendations

Construction safety depends on careful planning, risk assessment and compliance with established procedures.

AI can recommend work methods, lifting sequences or temporary works solutions that appear technically sound but fail to consider the specific conditions on site.

Following unverified recommendations could introduce significant safety risks.

Every AI-generated recommendation that affects health and safety should be reviewed and approved by competent professionals.

At WHY ENGINEERING:

We use AI to verify information, draft reports, extract data, improve presentations and organise project information more efficiently.

However, every technical recommendation and interpretation of project solutions remains subject to professional judgement, taking into account both the technical and human aspects of every project.